Skip to content

HALOWERK Agent Tools · tools.halowerk.com

Reads one conversion experiment two ways at once. Frequentist…

Two-proportion test, p value, confidence interval on the absolute lift, and the sample size the requested minimum detectable effect needs at the requested power. Bayesian: Beta-Binomial posteriors, probability the variant is better, expected loss either way, credible intervals - closed form, not sampled, so the numbers repeat exactly. looks_taken carries the adjusted threshold and the rate you actually ran at.

Ready now3 calls · 3 buyers in 30 days

Input

two_sided asks whether the groups differ, greater whether the variant is above the control.
Control
Baseline group. conversions may not exceed visitors.
How often the running test has been evaluated. Above 1 the threshold is adjusted and the real false positive rate is reported.
Relative effect the test should be able to find, e.g. 0.1 for ten percent above the control rate.
Probability of finding that effect if it is there.
Beta prior, successes. 1 with prior_beta 1 is uniform.
Beta prior, failures.
Variant
Group under test. conversions may not exceed visitors.

Sample response

Published by the provider. Your run returns a fresh response.

response9 keys
bayesian10 keys
expected_loss_choosing_control0.01169
expected_loss_choosing_variant0.000002
method"closed-form Beta-Binomial, no sampling"
probability_exacttrue
probability_variant_better0.997196
control3 keys
conversions241
rate0.05
visitors4820
deterministictrue
frequentist11 keys
absolute_lift0.013048
alpha0.05
alternative"two_sided"
confidence_level0.95
notenull
p_value0.0056071682
relative_lift0.26096
significanttrue
standard_error0.00471
z2.769911
peeking5 keys
adjusted_alpha0.01666667
correction"bonferroni"
effective_false_positive_rate_unadjusted0.142625
looks_taken3
significant_after_adjustmenttrue
sample_size6 keys
current_control4820
current_variant4790
minimum_detectable_effect_rel0.1
power0.8
required_per_group30245
sufficientfalse
statelesstrue
variant3 keys
conversions302
rate0.063048
visitors4790
verdict2 keys
decision"keep_running"
responseobject, 9 keys
Your agent can do this on its own
Add the MCP URL in Claude Code, Cursor or Codex. Your agent can run Tools within its spending limits, or ask for approval each time. Connect your agent

Connect your agent

Add one MCP URL and your agent can find and run this tool within its spending limits. Choose “Ask each time” if you want to approve every Call.

MCP URL

https://app.aecon.ai/mcp
Terminalbash
claude mcp add --transport http --scope user agentic-economy "https://app.aecon.ai/mcp"

In Claude Code, open /mcp, select agentic-economy, then choose Authenticate.

Track record

Paid calls, last 30 days
3
Buyers, last 30 days
3
Last paid call

Reported by the x402 directory.

A$0.0029 per call